| Message ID | 1421941147-9682-4-git-send-email-wens@csie.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Thu, Jan 22, 2015 at 11:39:05PM +0800, Chen-Yu Tsai wrote: > Since the dcdc regulator work frequency is a property of the regulators > DT sub-tree, there is no issue of naming collisions, so we can use the > standard "clock-frequency" binding for this. > > The driver sets a reasonable default value (same as the datasheet), and > only 1 dts has the old binding, with the default value. I don't see any benefit from this change - not only is this an incompatible change (which may affect out of tree users as well as in tree users) but it's not really doing anything with the clock API which I'd expect to be happening for a clock API property. What's the goal here?
diff --git a/drivers/regulator/axp20x-regulator.c b/drivers/regulator/axp20x-regulator.c index e4331f5e5d7d..990ac453abc8 100644 --- a/drivers/regulator/axp20x-regulator.c +++ b/drivers/regulator/axp20x-regulator.c @@ -186,8 +186,10 @@ static int axp20x_regulator_parse_dt(struct platform_device *pdev) if (!regulators) { dev_warn(&pdev->dev, "regulators node not found\n"); } else { - dcdcfreq = 1500; - of_property_read_u32(regulators, "x-powers,dcdc-freq", &dcdcfreq); + dcdcfreq = 1500000; + of_property_read_u32(regulators, "clock-frequency", &dcdcfreq); + /* axp20x_set_dcdc_freq() takes kHZ */ + dcdcfreq /= 1000; ret = axp20x_set_dcdc_freq(pdev, dcdcfreq); if (ret < 0) { dev_err(&pdev->dev, "Error setting dcdc frequency: %d\n", ret);
Since the dcdc regulator work frequency is a property of the regulators DT sub-tree, there is no issue of naming collisions, so we can use the standard "clock-frequency" binding for this. The driver sets a reasonable default value (same as the datasheet), and only 1 dts has the old binding, with the default value. Signed-off-by: Chen-Yu Tsai <wens@csie.org> --- drivers/regulator/axp20x-regulator.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)